Compton-edge-based Energy Calibration for Double-sided Silicon Strip Detectors in Compton Camera

Abstract

Double-sided silicon strip detectors (DSSDs) are considered as the most suitable detectors for the scatterer detectors in Compton cameras. The DSSDs show very high-energy and spatial resolutions; moreover, the DSSDs, which are made of silicon, a very low-Z material, show a minimal level of Doppler broadening, which is very important in Compton cameras especially for imaging low-energy photon sources. Currently, we are developing a novel type of Compton cameras, called double-scattering Compton imager (DOCI), which consists of two DSSDs as the scatterer detectors and a NaI(Tl) scintillation detector as the absorber detector. In this double-scattering-type Compton camera, the energy calibration of the scatter detectors (DSSDs) is crucial because the photon energy measured by one of the scatterer detectors essentially determines the scattering angle of the photon at the scatterer detector, directly affecting imaging resolution. The energy calibration of DSSDs is, however, very challenging for high-energy photons, e.g., > a few hundred keV. That is, the DSSDs are made of a low-Z material and the probability of photoelectric absorption is extremely low. Furthermore, the DSSDs are very thin (= 1.5 mm) and the probability of forming a full-energy peak from multiple interactions is also very low. Consequently, it is difficult to see a full-energy peak in the high-energy region, resulting in a failure of regular energy calibration. For the same reason, it is not possible to determine the energy resolution of the DSSDs for the high-energy region. In the present study, the energy calibration of the DSSDs was improved for the high-energy region by additionally considering the Compton edge in the energy spectrum. The present study used the Compton edge of 137Cs spectrum. The shape of the Compton edge strongly depends on the energy resolution of a DSSD and, therefore, a number of theoretical energy spectra were calculated, by Geant4 Monte Carlo detector simulations, changing the energy resolution from 0.5% to 20%. Then, each of the calculated energy spectra was compared with the measured spectrum in order to find the best matching spectrum. Finally, the information of the best matching spectrum was used to determine the location of the Compton-edge energy (= 477 keV) in the measured spectrum, which was then additionally used in subsequent energy calibration. To evaluate the effect of the improved energy calibration, the imaging resolution of the DOCI was determined for a 22Na point source with and without applying the Compton-edge-based energy calibration method. The imaging resolution before applying the calibration method was 9 mm FWHM for the 511 keV photons from the 22Na point source. The imaging resolution was improved by ~1 mm FWHM by the application of the calibration method. We believe that this Compton-edge-based calibration method is also very useful for other small and/or low-Z-material detectors, for which it is difficult to obtain full-energy peaks in the high-energy photon region.